Service Learning 1 Running Head: SERVICE LEARNING Can Service Learning and a College Climate of Service Lead to Increased Political Engagement After College? Nida Denson, M.A. Lori J. Vogelgesang, Ph.D. Victor Saenz, M.A. Higher Education Research Institute, UCLA 3005 Moore Hall, Box 951521 Los Angeles, CA 90095-1521 Paper presented at the Annual Meeting of the American Educational Research Association, Montréal, Canada, April 11-15, 2005
34
Embed
Service Learning 1 Running Head: SERVICE LEARNING · Service learning – the pedagogy of connecting academic learning with meaningful community service – represents one important
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Service Learning 1
Running Head: SERVICE LEARNING
Can Service Learning and a College Climate of Service
Lead to Increased Political Engagement After College?
Nida Denson, M.A.
Lori J. Vogelgesang, Ph.D.
Victor Saenz, M.A.
Higher Education Research Institute, UCLA 3005 Moore Hall, Box 951521 Los Angeles, CA 90095-1521
Paper presented at the Annual Meeting of the American Educational Research Association,
Montréal, Canada, April 11-15, 2005
Service Learning 2
Introduction
There is a growing national interest in strengthening the civic mission of higher education
(Boyte & Hollander, 1999; Erlich, 1999). This reflects concern with civic life in the United
States and a sense that Americans are “drawing back from involvements with community affairs
and politics.” (Skocpol & Fiorina, 1999, p.2). As a result, renewed attention is being given to
examining the role of educational institutions in their communities (Kellogg Commission, 1999)
and in preparing students to assume the responsibilities of citizenship in a democratic society
1944; Skocpol & Fiorina, 1999; Saltmarsh, 1996), but empirical work is less plentiful.
Furthermore, empirical studies that focus on civic engagement have not examined service
learning as a particular experience that might shape one’s propensity to be politically engaged
after college. Several studies have, however, examined other measures of civic and social
engagement (both attitudes and behaviors) in connection with service learning. This section will
first describe studies that focus on political engagement, and then share finding from research on
college students and service learning.
In a large national study of adults and civic volunteerism – including political
voluntarism – researchers Verba, Schlozman, and Brady (1995) examine both the motivations
and the capacities of different groups of Americans to participate in civic life. They conclude
that “voluntary activity in American politics suggests that the public’s voice is often loud,
sometimes clear, but rarely equal” (p. 509). The study finds that educational attainment is a
particularly important part of the social structure, is shaped by “circumstances of initial
privilege” and “has implications not only for the kinds of resources individuals accumulate but
also for the kinds of citizens they become. Educated citizens are much more likely to be
informed about politics and tolerant of unpopular opinions” (p. 514). Using data from the
Baccalaureate and Beyond Longitudinal Follow-up study, a study of college graduates (most in
their mid-twenties) also finds that educational attainment is a strong predictor of political
participation (voting, campaign volunteering, attending a political rally or meeting, contributing
Service Learning 5
money to a political campaign, and writing a letter to a public official) and civic participation
(Nie & Hillygus, 2001).
Both of these studies (Verba, Schlozman & Brady, 1995; Nie & Hillygus, 2001) suggest
that civic voluntary involvements reinforce political participation. Another study (a national
telephone survey), though, concludes that many adults choose either an electoral path to
involvement (20% of the sample) or a civic path (16% of the sample) (Keeter, Zukin, Andolina
& Jenkins, 2002). Only 16% of the sample reported being active in both realms. These
researchers find that over half of young people (ages 15-25) are disengaged, but they also find
that with help, young people are more likely to be engaged. “They respond to school-based
initiatives, at least in the short-run, as well as to other invitations to involvement” (p. 2). The
Keeter et. al (2002) study also explores a new area of political voice – consumer activism.
Buying, or choosing not to buy, a product or service because of the political values of a
company, is widespread among all age groups (except those born before 1946), and is most
prevalent among those who are better educated and more affluent. Consumer activism is most
prevalent among those who are already political and civic activists.
Service Learning Research
Although research findings are not entirely consistent, there are enough connections
between civic voluntarism and political engagement to warrant further investigation. In recent
years, the pedagogy of service learning has come to be seen as one way to get students
‘practicing’ civic engagement at a young age. Will service lead to social action? As Craig
Rimmerman sums it up:
The hope on the part of many service organizers is that students who participate in service activities will begin to ask why tragedies such as illiteracy, hunger, and homelessness even exist. Thus such students, many of whom are apolitical, will begin to develop a social consciousness.” (1997, p. 103).
Service Learning 6
But ideas of strengthening students’ sense of civic responsibility, although mentioned in
definitions of service learning, are not necessarily at the forefront of any faculty member’s list of
course outcomes as he or she designs a course. Some faculty members and other service
learning proponents speak of the potential of the pedagogy as a social justice mechanism, but
there is not agreement within the service learning field that social justice ought to be an intended
outcome of service learning participation (Zlotkowski, 1996; Marullo & Edwards, 2000); many
instructors see service learning simply as a good way to teach academic course content. And
indeed, large studies have documented that service learning has the potential to improve learning
outcomes, particularly when connections between the service experience and the classroom
experience are strong (Astin, Vogelgesang, Ikeda & Yee, 2000; Eyler & Giles, 1999; Gray et al.,
1999).
Although civic engagement outcomes may or may not be built into service learning
courses, Eyler and Giles argue persuasively that service learning and higher education “need to
pay attention to the problem-solving capacities of college graduates in order to sustain lifelong
constructive involvement in the community” (1999, p. 155). In particular, their study explores
five elements of citizenship: Values, knowledge, skills, efficacy and commitment. Service
learning can enhance each element, but the authors note that the ultimate test of the final
element, commitment, is behavior. Other empirical work has supported Eyler and Giles’
findings that service learning strengthens civic values, skills and efficacy (Vogelgesang & Astin,
2000). However, in these and other studies that are limited to the college years, commitment to
values and ideas serves as a proxy for (longer-term) post-college behavior. This study seeks to
explore whether these values and beliefs translate into strengthened civic values, attitudes, and
importantly, behaviors in the post-college years.
Service Learning 7
Theoretical Framework
This study focuses on one aspect of civic participation: political engagement in the post-
college years. Although other studies have found that level of education is positively correlated
with voting and some forms of involvement (Nie & Hillygus, 2001; Verba, Schlozman & Brady,
1995), this study takes a different perspective in that it seeks to understand whether a specific
activity – service learning – performed during the college years impacts post-college political
engagement, and secondly whether a climate of service on a college campus can strengthen the
impact of service learning participation on political engagement.
We define the climate for service as the aggregate level of students’ service participation
at the respective institutions of the study’s respondents. (Note: Because the cohort examined in
this study entered college in 1994, when service learning was not as common as it is today, we
chose to focus on the effects of the college environment for generic service). As a result, the
theoretical framework/perspective for the study is informed by an understanding of the impact of
peer groups, and the extent to which college peer groups have post-college effects. In his work
examining how college affects students, Astin (1993) posits a theory of peer group influence,
which theorizes that students will shape their behavior to the norms and expectations of group
members, and thus “students tend to become more like their peers” over the course of their
college years (p. 402). Other researchers have also suggested that the peer group has a powerful
effect on individual student activities and beliefs (Chickering, 1969; Feldman & Newcomb,
1969). Accordingly, we examine not only whether respondents become more like their peers,
but also how their college peer environments shape post-college outcomes in ways that other
environments might not. Ultimately, this longitudinal study examines how a service learning
Service Learning 8
experience and a climate of volunteerism in general affect respondents’ political engagement in
the post-college years.
Method
The primary focus of this study is to assess the effects of service learning during college
on post-college political engagement. Further, this study explores the impact of peer average
levels of volunteerism/service on respondents’ political engagement in the post-college years.
The following section offers analytic details of the study, including a description of the data,
sample, variables, and analysis.
Sample
The data for this study was collected as part of the Cooperative Institutional Research
Program (CIRP), which is sponsored by the Higher Education Research Institute (HERI) at the
University of California, Los Angeles. This study was also supported by a three-year grant from
the Atlantic Philanthropies U.S.A., Inc. The data for the first time point was collected when this
cohort entered college in 1994, using the Student Information Form (SIF) which is designed as a
pre-test for longitudinal assessments of the impact of college on students. The paper and pencil
instrument surveys incoming freshmen students about their activities during high school, as well
as their values, beliefs and attitudes. The data for the second time point comes from the 1998
College Student Survey (CSS). Students were administered this second survey in 1998, at the
end of their fourth year in college. The 1998 CSS follow-up sample was chosen from the
original students who completed the 1994 SIF, and consists of approximately 20,000 students.
The data for the third time point comes from the Post-College Follow-up Survey (PCFS),
administered in 2004. The PCFS follows the 1994 entering cohort and covers activities such as
participation in the political process and specific ways in which individuals are involved in
Service Learning 9
serving their communities (professional organizations, non-profit work, issue-oriented
involvement, etc.), as well as beliefs and values. Since the 2004 survey collects rich data on
forms of political and civic participation since leaving college, it provides a more sophisticated
dependent measure, which reflects not only voting and discussing politics, but a variety of ways
in which individuals move to influence the political structure.
The initial sample for this study consisted of 8,434 respondents from 229 institutions who
completed the 1994 CIRP, the 1998 CSS, and the 2004 PCFS. The final sample used for this
study was different from the full longitudinal sample just described. First, we excluded students
with missing data on race. We also excluded students who had marked “American Indian” for
race (too small sample size). We then excluded institutions with missing data on selectivity,
size, and institutions with less than 20 cases. The final sample consisted of 7,248 students nested
within 150 institutions who were surveyed upon entering college in 1994, followed-up in 1998 as
graduating seniors, and then followed-up again in 2004 ten years after entering college. The
final sample of students consisted of 2,403 (33.2%) males and 4,845 (66.8%) females. Of these
students, 6,620 (91.3%) were White, 265 (3.7%) were Asian American, 185 (2.6%) were African
American, and 178 (2.5%) were Latino/a. The final sample of institutions consisted of 33
(22.0%) universities and 117 (78.0%) four-year colleges. Furthermore, 27 (18.0%) of the
institutions were public, while 123 (82.0%) were private.
Political Engagement Outcome
The political engagement factor was comprised of 17 PCFS items that relate to various
forms of political engagement. Factor analysis was used to confirm the general political
engagement factor, and had an alpha reliability of 0.89 with factor loadings of at least 0.46 or
greater. Factor loadings of each of the individual items are in parentheses.
Service Learning 10
• Please indicate the importance to you personally of each of the following: Influencing the political structure (.70) Influencing social values (.50) Keeping up to date with political affairs (.64)
• Please indicate if you have performed any of the following since leaving college: Donated money to a political candidate or cause (.64) Expressed your opinion on a community or political issue by contacting or
visiting a public official (.67) Worked with a political group or official (.68) Worn a campaign button, put a sticker on your car, or placed a sign in front of
your house supporting an issue or candidate (.70) Worked as a canvasser going door to door for a political candidates or a cause
(.50) Used on-line communication with family and friends to raise awareness about
social and political issues (.65) Expressed your opinion on a community or political issue by signing a written or
email petition (.64) Bought a certain product or service because you liked the social or political values
of the company (.63) Not bought something or boycotted it because of the social or political values of
the company (.61) Voted in a national election (.46) Voted in a state/local election (.49)
• Since leaving college, how often have you participated in community service/volunteer work through the following organizations?
A political organization (e.g., political party, campaign, etc.) (.69) • For the activities listed below, please indicate how often have you engaged in each during
the past year: Participated in protests/demonstrations/rallies (.58) Discussed politics (.62)
Independent Variables
The principal independent variables of interest are all 1998 variables that relate to
service: 1) a dummy variable indicating whether a student participated in any service learning
during college – meaning that students who participated in both service learning and
volunteering were also included in this group (a student-level variable), 2) a dummy variable
indicating whether a student volunteered only during college (a student-level variable), and 3)
Service Learning 11
peer average levels of volunteerism which represents the percentage of all respondents for that
institution who reported volunteering during college (an institution-level variable).
In testing the effects of service learning and volunteerism at multiple levels on student
outcomes, other key variables were also included in the analyses to account for differences in
precollege characteristics and college socialization, and to control for the effects of certain
critical institutional characteristics (see Appendix A). These variables were selected in order to
rule out alternative explanations for findings.
In general, the suggested guidelines for regression analysis is at least 10 observations for
each predictor, however, the corresponding rules for hierarchical models are somewhat more
complex due to the statistical consideration of multiple levels. That is, there should also be at
least 10 institutions per institution-level predictor in the model (Raudenbush & Bryk, 2002).
Thus, due to these methodological constraints we were very deliberate in deciding which
variables to include in our final analyses, and where possible, created composites as a way to
reduce the number of variables (e.g., faculty support was a factor comprised of seven items).
The sets of variables included in our final analyses are presented next.
Student-level. The first set of student-level variables consisted of a freshmen pretest for
political engagement as well as variables representing students’ precollege characteristics. The
freshman pretest factor associated with the outcome measure of overall political engagement,
consisted of the following four items from the 1994 SIF: discussed politics, importance of
influencing the political structure, importance of influencing social values, and importance of
keeping up to date with political affairs. These 1994 SIF variables were chosen because they
exactly mirror some of the items that make up the 2004 outcome. Since the 2004 survey was
created specifically to assess forms of political and civic participation six years after college,
Service Learning 12
while the 1994 SIF and 1998 CSS surveys were designed for longitudinal assessments of the
impact of college on students, there were only these four specific items on the 1994 SIF that
exactly mirrored items on the political engagement factor on the 2004 survey. The respondents’
pre-college characteristics consisted of such variables as gender, ethnicity (Whites, Asian
Americans, African Americans, and Latino/as), parental education (composite of father’s and
mother’s education), high school volunteering, high school GPA, and 1994 political orientation.
A second set of variables controlled for individual college experiences such as college
major, faculty support, and other college activities. Since we had to carefully limit the number
of variables to be included in the analyses, it was not possible to include all college majors
individually (e.g., on the CSS survey there are 44 possible majors listed). Instead, a dichotomous
major variable was created (1=Arts/Humanities/History/Political Science; 0=All other majors).
These four specific major categories were combined because the students in these majors were
more likely to be politically engaged post-college. The faculty support variable was a composite
of seven items (α=.83). The first item asked respondents in 1998 how often they felt faculty
took a personal interest in their progress (1=not at all to 3=frequently). The next six items asked
how often professors at their current college provided them with the following (1=not at all to
3=frequently): advice and guidance about your educational program, respect (treated you like a
colleague/peer), emotional support and encouragement, honest feedback about you skills and
abilities, intellectual challenge and stimulation, and an opportunity to discuss coursework outside
of class.
Two other variables in this set consisted of whether a student joined a fraternity or
sorority and whether a student worked full-time (both items are dichotomous: 0=not marked;
1=marked). There was a composite variable indicating frequency of involvement in
Service Learning 13
curricular/co-curricular diversity activities: enrolled in an ethnic studies course, enrolled in a
women’s studies course, attended a racial/cultural awareness workshop, and joined a
racial/ethnic student organization (0=not marked; 1=marked). And, another composite variable
from the 1998 survey consisted of three items asking respondents whether or not they had
participated in the following college activities: student government, in honors/advanced courses,
and in leadership training (0=not marked; 1=marked). It should be noted that these last two
composite measures serve to capture only the quantity (not quality) of certain types of curricular
involvement and campus involvement, respectively. While these variables are not of primary
substantive interest, they were included in the analyses to reduce the risk of overestimating the
effects of service and service learning.
Institution-level. The institution-level control variables included institutional control
(public/private), enrollment size (number of undergraduate FTE), level of selectivity, and
percentage of underrepresented minority students at the institution. These variables were
included because they are well-known structural differences that shape student experiences in
higher education and also enable us to control for sample biases. We also included at this level
the aggregate measures of the student-level variables for all the respondents within each
institution so that we can better differentiate student vs. institution-level effects and rule out
other potential unique culture/climate effects. Appendix A lists descriptive statistics for all the
variables included in the analyses.
Analytic Approach
Our main focus, as noted earlier, is to examine the effects of individual participation in
service learning and peer average levels of volunteerism in college on the outcome of political
engagement ten years after college entry. In order to differentiate the effects of service learning
Service Learning 14
and generic volunteerism, both of these non-overlapping variables were included in the analyses
(Note: Those students who participated in both service learning and generic volunteerism were
included in the service learning group). Furthermore, because of the multilevel nature of the
contextual research question – the influence of peer average levels of volunteerism in college on
political engagement post-college – this type of analysis could only be accomplished using a
multilevel methodology such as HLM.
The problems of neglecting the hierarchical or nested nature of the data gathered by using
a single-level statistical model have been acknowledged and addressed by a number of